Optimizing mobile multimedia content delivery

With the advent of mobile Internet the amount of time people spend with multimedia applications in the mobile environment is surging and demand for high quality multimedia data over the Internet in the mobile environment is growing rapidly. However the mobile environment is significantly more unfriendly than the wired environment for multimedia applications in many ways. Network resources are limited and the condition is harder to predict. Also multimedia applications are generally delay intolerant and bandwidth demanding, and with users moving, their demand could be much more dynamic and harder to anticipate. Due to such reasons many existing mobile multimedia applications show unsatisfactory performance in the mobile environment. We target three multimedia content delivery applications and optimize with limited and unpredictable network conditions typical in the mobile Internet environment. Vehicular networks have emerged from the strong desire to communicate on the move. We explore the potential of supporting high-bandwidth applications such as video streaming in vehicular networks. Challenges include limited and expensive cellular network, etc. Internet video conferencing has become popular over the past few years, but supporting high-quality large video conferences at a low cost remains a significant challenge due to stringent performance requirements, limited and heterogeneous client. We develop a simple yet effective Valiant multicast routing to select application-layer routes and adapt streaming rates according to dynamically changing network condition in a swift and lightweight way enough to be implemented on mobile devices. Bitrate adaptive video streaming is rapidly gaining popularity. However recent measurements show weaknesses in bitrate selection strategies implemented in today's streaming players especially in the mobile environment. We propose a novel rate adaptation scheme that classifies the network condition into stable and unstable periods and optimizes video quality with different strategies based on the classification.